A large literature documents the short-term benefits of stimulant medication for the treatment of ADHD. However, a regular increase in dose is typically needed to maintain full effect, and there is a little support in the literature for long-erm benefit even when medication is consistently taken over a decade. We consider this unexplained loss of effectiveness to be one of the most important issues in child psychopharmacology. Tolerance, as defined as a state of adaptation in which exposure to a drug results in a decrease of the drug's effects over time, is one possible explanation for this observed lack of sustained effect. Concerns about possible tolerance to therapeutic effects of stimulants date back 25 years. Tolerance is a potentially treatable phenomenon and has been the area of much research in pain medicine. However, there have been no randomized trials to evaluate the presence of tolerance to CNS stimulants in ADHD populations or evaluations of how to manage it. To address these gaps in the literature, we propose a two phase translational study: (1) an innovative evaluation of short-term tolerance using the most well established protocol for measuring controlled stimulant effects - the analog classroom - in a within subjects design during the summer; and (2) a school-year follow-up of the same participants to evaluate (a) individual differences in the development of tolerance over sustained treatment, and (b) if weekend drug holidays mitigate the need for dose increases. Specifically, 250 youth with ADHD will be stabilized on an optimal dose of OROS MPH based on performance in the summer analog classroom setting. Participants will be tested over 3-weeks of exposure to both OROS MPH and placebo in a crossover design. Within-subject drug-placebo differences in performance on a timed math task will be compared over the three weeks of exposure. Individual tolerance estimates will be computed for each subject (AIM 1). In phase 2, all subjects will begin the school year on their optimal summer OROS MPH dose and be randomized to 7 (continuous dosing) or 5 day a week dosing (weekend holidays). Study physicians will prescribe stimulant medication for the duration of the school year. Subjects will be monitored monthly using standardized measures of response. When a participant exhibits suboptimal symptom control (possible tolerance) determined by predetermined thresholds derived from the MTA, dose will be increased. Group differences in dosage needs and symptom severity will be examined (AIM 2). The ability of the tolerance estimates from the analog classroom to predict subsequent dosing patterns for individual subjects will be examined (AIM 3) to ascertain if laboratory observed tolerance predicts real world dosing.